Vessel Geometry Interacts with Red Blood Cell Stiffness to Promote Endothelial Dysfunction in Sickle Cell Disease

Introduction: Sickle cell disease (SCD) is a genetic blood disorder in which red blood cell (RBC) stiffness is abnormally increased. In addition, chronic endothelial dysfunction, or vasculopathy is another aspect of SCD that involves RBC–endothelial cell interactions, although the underlying mechanisms remain poorly understood. Recent experimental work shows that stiffer RBCs marginate towards the blood vessel walls under physiologic flow conditions due to cell-cell collisions. However, little research has focused on the mechanical interactions between flowing stiff RBCs and endothelium in SCD that are not in the context of vascular occlusion in deoxygenated conditions. We propose that stiff, sickled RBCs in SCD patients constantly interact with the endothelium due to this stiffness-mediated margination, and that this interaction constitutes a purely mechanical cause of endothelial cell dysfunction. Furthermore, we hypothesize that the blood vessel geometry, which controls blood flow patterns and shear stress cellular, will mediate this mechanically-based endothelial dysfunction and may be an important aspect in the development of this vasculopathy. However, an adequate experimental model to test this hypothesis does not exist. To that end, we developed a simple “do-it-yourself” (DIY) perfusable vasculature model that incorporates a confluent endothelial cell monolayer along the channel lumen and recapitulates complex vascular geometries such as curvature.

Materials and Methods: To fabricate the DIY endothelialized vasculature model, a strand of 500um diameter PMMA optical fiber was cast and cured in PDMS. The optical fiber was removed, leaving behind a channel that was then cultured with human aortic endothelial cells (HAECs). Bends were introduced into the fibers to create curved geometries. To test the effect of stiff RBCs on the endothelium, suspensions of RBCs from SCD patients were infused into these endothelialized devices, and assessed for endothelial dysfunction via immunostaining for VCAM-1 and E-selectin, known markers of endothelial inflammation. These were then compared to devices infused with control RBCs. To decouple the potential biological causes of endothelial dysfunction in SCD (e.g., adhesion, hemolytic byproducts) from purely physical causes, normal RBCs were dehydrated with nystatin concentrations known to increase the RBC stiffness to similar levels of SCD.

Results and Discussion: These DIY vasculature models recapitulate in vivo microvasculature and can be cultured with human aortic endothelial cells (HAECs). (Fig 1.A,B). Simulations show an acute and localized shear rate variability at the site of curvature (Fig 1C). HAECs exposed to SCD RBCs and nystatin-stiffened RBCs perfused at flow rates of 100µL/min exhibited increased VCAM-1 and E-selectin upregulation in the curved regions of the vessel with little effect upstream or downstream of that region (Fig. 2). More specifically, SCD patients show increased endothelial inflammation along the outside wall of the bend (Fig 2). This is an interesting result as the regions of high wall shear stress associated with endothelial dysfunction occur along the outside wall of curved vessels. We speculate that the endothelial inflammation occurring in our system is related to increased collisions with the stiff SCD RBCs that occurs when the stiff RBCs marginate preferentially to the outer wall due to the inertial effects created by fluid flow around a bend. HAECs exposed to RBCs artificially stiffened with nystatin, however, showed increased diffuse VCAM-1 and E-selectin expression throughout the entire region of the curvature compared to healthy and SCD RBCs, potentially due to higher degrees of RBC margination compared to SCD (Fig 2). Overall, these results indicate that the mechanical interactions between stiff RBCs and the endothelium, as well as vascular geometry, plays a role in SCD vasculopathy. Additionally, studies investigating systematically quantifying the effect of varying degrees of vessel curvature on endothelial dysfunction.

Conclusion: These results provide new explanations for the complex causes of endothelial dysfunction in SCD by relating the mechanical properties of RBCs as well as the vessel geometry to endothelial cell inflammation. Particularly, these studies have profound implications for understanding stroke in SCD, due to the tortuosity of the cerebral vasculature.